Open dansteren opened 2 years ago
dansteren
commented
2 years ago It could also be good to have an official time-period for voting so that we aren't waiting too long before proceeding.
Can't quite tell if it's related to this, but wanted to mention it.
legendface66
commented
2 years ago I suggest instead of voting we produce an opinion, save it as a docusign and working group participants sign it off. If working group participants don't want to sign they can produce an alternative opinion and circulate it for signatures. This is better than voting for the following reasons:
It will help identify participants who are actually participating. It creates an opportunity for debate and forces dissenters to explain their dissent. The difference between the two opinions will help us identify key issues that we need to work on as a community. It also provides a clear chain of documentation.
[Edit] It is not clear to me that we need to achieve consensus. Motion proposal voting in the NNS can provide that. What we need is effectively explained and well-reasoned/evidenced opinions.
lastmjs
commented
2 years ago I'm starting to think we should go with a rough consensus. I propose that pull requests be the main means of producing the official opinions of the working group. If there are no substantial down votes on a pull request after a sufficient period of time, then I think it should be merged in. We can look at the ratio of up votes to down votes on pull requests, it should be obvious when something is controversial. When something is controversial, then it should not be merged in.
dietersommer
commented
2 years ago Agreeing with @lastmjs. In the Ledger and Tokenization Working Group we are discussing using the concept of rough consensus as the basic means for getting agreement within the WG. We are discussing whether WG-level votes make sense or whether they can be eliminated or only used in case rough consensus does not lead to a result. But in this case, the vote will not lead to consensus and thus be suboptimal.
legendface66
commented
2 years ago I don't understand the need for consensus in a working group. We are not making decisions, we are doing research and providing opinions. I suggest we shelve this conversation at least until some work has been completed and is ready for publication. Then if:
We will have encountered a blocker requiring consensus regarding the seal of approval. This blocker will need to be cleared. However, right now it does not appear to me that this is required.
dietersommer
commented
2 years ago Let me give you some thoughts on this. A technical working group wants to have consensus in the sense that not too many people have objections to a proposal the group wants to subject to an NNS note. The group is making a standards proposal and not having, at least rough, consensus on it means that the proposal may have room for (technical) improvement, i.e., is suboptimal at the time. This assumes that objections are at a technical level. With a process of having rough consensus, we can increase the technical quality of proposals. Have a look at how IETF is doing it: https://www.rfc-editor.org/rfc/rfc7282
dietersommer
commented
2 years ago We have a draft for some governing principles for IC working groups (ICRC-0). Would be great if people from the governance WG could have a look and comment and help improve! https://forum.dfinity.org/t/ledger-tokenization-working-group-update/15298/42
As PRs are created to propose official adoptions of the working group, we will need a way to vote on them and a specific metric by which we can determine if a proposal has passed.
My baseline assumption is that we will express our votes with :+1: and :-1: reactions to a PR. However, how will we know if a proposal has passed? In other working groups a large majority usually led to a measure passing, while a roughly 50/50 vote would signify additional discussion was needed.
In order for everyone to feel good about measures passing/failing I think it'd be helpful if we determined a vote percentage that we need to consider something passed. It could be 100% could be >50%. Whatever it is, we should all be in agreement on it though.